Modern computational mathematics increasingly relies on simultaneous multi-way tensor arrays exceeding petabyte scales. Representing interconnected mathematical variables yields sparse higher-order tensors whose dimensions easily surpass $10^5 \times 10^5 \times 10^4$. Traditional centralized CANDECOMP/PARAFAC (CP) and Tucker factorization routines exhaust memory bandwidth on single compute nodes, stalling downstream computational discovery.
In this work, we propose Distributed Tucker-CP (DT-CP), an asynchronous, communication-avoiding tensor decomposition framework tailored for high-performance computing clusters with heterogeneous GPU accelerators. By interleaving tensor-times-matrix (TTM) contraction with lock-free parameter replication, DT-CP reduces inter-node synchronization stalls by up to 88%.
2. Mathematical Formulation & Architecture
Given an $N$-th order tensor $\mathcal{X} \in \mathbb{R}^{I_1 \times I_2 \times \dots \times I_N}$, our objective is to compute low-rank factor matrices $\mathbf{A}^{(n)} \in \mathbb{R}^{I_n \times R_n}$ and a dense core tensor $\mathcal{G} \in \mathbb{R}^{R_1 \times R_2 \times \dots \times R_N}$ minimizing the Frobenius norm error:
We partition the input tensor along spatial Cartesian sub-grids mapped onto a 3D torus interconnect topology. Each GPU worker executes local sparse tensor contractions using custom CUDA warp-shuffle primitives, buffering intermediate matricized gradients in asynchronous ring buffers.
3. Empirical Performance & Numerical Benchmarks
We evaluated DT-CP against baseline distributed frameworks on high-performance supercomputing nodes. Time-to-convergence for a rank-(50, 50, 20) decomposition dropped from 272.4 minutes with MPI-Tensor to 18.4 minutes with DT-CP, delivering a 14.8x net speedup while reducing peak host RAM allocation by 64%.
4. Conclusions & Future Horizons
DT-CP establishes a scalable, mathematically rigorous foundation for extreme-scale tensor analytics in applied mathematics and computational physics.
Funding & Support
National Science Foundation (Grant OAC-2311904) and Mathematical Sciences Research Council.
Competing Interests
The authors declare no competing financial or non-financial interests.
Author Contributions (CRediT Taxonomy)
Author User: Conceptualization, Methodology, Software, Writing – original draft
Prof. Sarah Chen: Validation, Supervision, Writing – review & editing
System Administrator: Investigation, Formal analysis, Funding acquisition
Figure 1
Schematic of the Distributed Tucker-CP (DT-CP) asynchronous pipeline. Tensor slices are distributed across a 3D GPU torus with lock-free ring-buffered gradient exchanges.
Figure 2
Weak and strong scalability benchmarks comparing DT-CP against baseline MPI-Tensor solvers across 16 to 256 compute nodes.
[1]
Kolda, T. G., & Bader, B. W. (2009). Tensor decompositions and applications. SIAM Review, 51(3), 455-500.
Vance, E., & Miller, D. (2025). Asynchronous lock-free coordination for extreme-scale scientific computing. Journal of Supercomputing Frontiers, 12(2), 88-104.
Usage metrics conform to COUNTER Release 5 standard reporting protocols, filtering out automated search crawlers and rapid consecutive refresh requests.
User, A., Chen, P. S., & Administrator, S. (2026). Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
APA 7th Edition
More Citation Formats
Click any format to view & copy:
User, A., Chen, P. S., and Administrator, S. 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics. 1, 2 (2026), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
ACM Format
User, A.; Chen, P. S.; Administrator, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM 2026, 1 (2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
ACS Format
User, A., Chen, P. S., & Administrator, S. (2026). Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
APA Format
USER, A.; CHEN, P. S.; ADMINISTRATOR, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, v. 1, n. 2, p. 101-118, 2026. Disponível em: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
ABNT Format
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Chicago Format
User, A., Chen, P. S., and Administrator, S., 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), pp.101-118. Available at: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
Harvard Format
A. User, P. S. Chen, and S. Administrator, "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping," IJPACM, vol. 1, no. 2, pp. 101-118, 2026, doi: 10.5555/ijpacm.2026.1.2.01.
IEEE Format
User, Author, et al. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics, vol. 1, no. 2, 2026, pp. 101-118, https://doi.org/10.5555/ijpacm.2026.1.2.01.
MLA Format
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Turabian Format
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118.
Vancouver Format
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118. doi:10.5555/ijpacm.2026.1.2.01
Select from 11 academic styles or export structured references.
ACM Format
User, A., Chen, P. S., and Administrator, S. 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics. 1, 2 (2026), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
ACS Format
User, A.; Chen, P. S.; Administrator, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM 2026, 1 (2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
APA (7th) Format
User, A., Chen, P. S., & Administrator, S. (2026). Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
ABNT Format
USER, A.; CHEN, P. S.; ADMINISTRATOR, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, v. 1, n. 2, p. 101-118, 2026. Disponível em: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
Chicago Format
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Harvard Format
User, A., Chen, P. S., and Administrator, S., 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), pp.101-118. Available at: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
IEEE Format
A. User, P. S. Chen, and S. Administrator, "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping," IJPACM, vol. 1, no. 2, pp. 101-118, 2026, doi: 10.5555/ijpacm.2026.1.2.01.
MLA (9th) Format
User, Author, et al. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics, vol. 1, no. 2, 2026, pp. 101-118, https://doi.org/10.5555/ijpacm.2026.1.2.01.
Turabian Format
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Vancouver Format
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118.
AMA (11th) Format
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118. doi:10.5555/ijpacm.2026.1.2.01
@article{user2026_7155,
title = {{Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping}},
author = {Author User and Prof. Sarah Chen and System Administrator},
journal = {International Journal of Pure, Applied and Computational Mathematics},
volume = {1},
number = {2},
pages = {101-118},
year = {2026},
publisher = {Academic Mathematical Publishing House},
doi = {10.5555/ijpacm.2026.1.2.01},
url = {https://doi.org/10.5555/ijpacm.2026.1.2.01},
issn = {2348-0084}
}
TY - JOUR
TI - Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping
T2 - High-Performance Asynchronous Factorization on Heterogeneous GPU Clusters
AU - Author User
AU - Prof. Sarah Chen
AU - System Administrator
JO - International Journal of Pure, Applied and Computational Mathematics
VL - 1
IS - 2
SP - 101
EP - 118
PY - 2026
DO - 10.5555/ijpacm.2026.1.2.01
UR - https://doi.org/10.5555/ijpacm.2026.1.2.01
PB - Academic Mathematical Publishing House
SN - 2348-0084
AB - High-throughput multi-omics sequencing generates multi-way tensor arrays exceeding petabyte scales, creating urgent computational bottlenecks for clinical discovery. We introduce Distributed Tucker-CP (DT-CP), an asynchronous lock-free tensor factorization algorithm optimized for CUDA/ROCm memory hierarchies with adaptive communication pipelining. On a 256-node GPU cluster, DT-CP demonstrates a 14.8x acceleration over state-of-the-art MPI-Tensor frameworks, processing 4.2 billion patient feature interactions in 18.4 minutes while maintaining 99.4% spectral accuracy. DT-CP unlocks real-time multi-omics phenotyping in clinical genomics pipelines, providing an open-source mathematical infrastructure for precision medicine.
KW - Tensor Decomposition
KW - Numerical Linear Algebra
KW - High-Performance Computing
KW - GPU Acceleration
KW - Asynchronous Algorithms
ER -
All 11 supported citation standards for this article are shown below. Click "Copy" next to any style to copy it directly to your clipboard.
ACM
User, A., Chen, P. S., and Administrator, S. 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics. 1, 2 (2026), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
ACS
User, A.; Chen, P. S.; Administrator, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM 2026, 1 (2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
APA 7th
User, A., Chen, P. S., & Administrator, S. (2026). Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01
ABNT
USER, A.; CHEN, P. S.; ADMINISTRATOR, S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, v. 1, n. 2, p. 101-118, 2026. Disponível em: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
Chicago
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Harvard
User, A., Chen, P. S., and Administrator, S., 2026. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. International Journal of Pure, Applied and Computational Mathematics, 1(2), pp.101-118. Available at: <https://doi.org/10.5555/ijpacm.2026.1.2.01>.
IEEE
A. User, P. S. Chen, and S. Administrator, "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping," IJPACM, vol. 1, no. 2, pp. 101-118, 2026, doi: 10.5555/ijpacm.2026.1.2.01.
MLA 9th
User, Author, et al. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics, vol. 1, no. 2, 2026, pp. 101-118, https://doi.org/10.5555/ijpacm.2026.1.2.01.
Turabian
User, Author, Chen, Prof. Sarah, and Administrator, System. "Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping." International Journal of Pure, Applied and Computational Mathematics 1, no. 2 (2026): 101-118. https://doi.org/10.5555/ijpacm.2026.1.2.01.
Vancouver
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118.
AMA 11th
User A, Chen PS, Administrator S. Scalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping. IJPACM. 2026;1(2):101-118. doi:10.5555/ijpacm.2026.1.2.01
DOI: 10.5555/ijpacm.2026.1.2.01
PDF GalleyScalable Distributed Tensor Decomposition for Multi-Omics Clinical Phenotyping
Modern scientific workflows necessitate full algorithmic transparency and verifiable computation. In this investigation, we establish the foundational principles governing high-throughput data representations.
Continuous parameter spaces are mapped using adaptive loss functions with explicit bound convergence.
2. Experimental Framework
Empirical validation was performed on standardized benchmarking clusters. Communication profiles and latency bottlenecks were captured at sub-millisecond granularity.
All source algorithms are fully archived under open-source licenses for independent replication.
ScholarlyPress Galley Typeset • June 15, 2026Page 1 of 18
Document Size: 2.8 MB • Verified SHA-256 Checksum
NISO Z39.96
JATS 1.2 Archival XML
Standardized Journal Article Tag Suite (JATS) schema for automated deposits to PubMed Central, Crossref, and institutional repositories.
The manuscript has been converted into full-text academic HTML, assigned to the current journal volume, and is now live and indexable worldwide without authentication.
Be the first to read newly accepted peer-reviewed manuscripts. Enter your email to receive immediate full-text notifications whenever a new scholarly article is officially published by the journal.
✓
Subscription Confirmed!
You're all set. We will send you notification and download links for every new paper published.